Using the finite domain constraint solver of GNU Prolog version 1.2.16 on a Mac OS X 10.4.4 PPC machine the behaviour of fd_labeling/1 seems to be not correct with respect to constraints set up prior using fd_labeling/1.
At least the behaviour is different if the sequence of goals is different
with respect to fd_labeling. The following sequence of goals generates the expected answer: | ?- fd_domain([X,Y],1,3),fd_labeling([X,Y]),fd_all_different([X,Y]). X = 1 Y = 2 ? a X = 1 Y = 3 X = 2 Y = 1 X = 2 Y = 3 X = 3 Y = 1 X = 3 Y = 2 (1 ms) no | ?- Changing the sequence of the two last goals not all possible combinations of values for X and Y are generated. Even more fatal: one combination is generated which is forbidden by the constraint. | ?- fd_domain([X,Y],1,3),fd_all_different([X,Y]),fd_labeling([X,Y]). X = 1 Y = 2 ? a X = 1 Y = 3 X = 2 Y = 3 X = 3 Y = 3 (1 ms) yes | ?- Markus Morawitz [EMAIL PROTECTED] _______________________________________________ Bug-prolog mailing list [email protected] http://lists.gnu.org/mailman/listinfo/bug-prolog
